Project Summary Over 100 million Americans will be over age 65 years by 2060 (representing 25% of the U.S. population), with an increasing proportion of African-Americans. Understanding which Americans exhibit ?exceptionally healthy aging??an older life free of illness, disability, or functional limitation?is central to efforts to improve quality of life and resource utilization. Current investigations on ?healthy aging? focus on systemic inflammatory markers (e.g., interleukin-6) or composite indices of organ/metabolic function (the ?healthy aging index,? HAI), both lim- ited by a (1) lack of sensitivity for early changes in metabolism; (2) lack of specificity for molecular pathways involved in these changes; (3) absence of investigation early in adulthood (when ?aging? may be reversible); (4) few studies of African-Americans, at high-risk for age-related morbidity. In preparation for this application, we applied metabolite profiling in a pilot study of nearly 300 African-American males in Health ABC, a biracial NIA study of elderly Caucasians and African-Americans, identifying metabolites associated with HAI (oxidative stress, nitric oxide signaling, gut microbial metabolism), some of which may be unique to African-Americans. Here, we hypothesize that circulating metabolites associated with multi-dimensional aging phenotypes and outcomes in older adults will define pathways of ?exceptional healthy aging? that may be race-specific and dysregulated in youth at risk for ?premature? aging. To address this hypothesis, we will harness two well- characterized, biracial American cohorts that span life (elderly: age 733 N=1312; young: CARDIA, age 324 N=2376; R01-HL136541). In Aim 1, we define metabolic pathways implicated in race-specific healthy aging, based on association between metabolites and validated age-related phenotypes heterogeneous by race (physical, neurocognitive function, vascular structure, body composition). We will assess determinants of iden- tified pathways, including dietary quality, physical activity, and psychosocial stress (poorer in African- Americans). In Aim 2, we identify metabolites related to ?exceptionally healthy aging? in the elderly (freedom from disability, dementia, cancer, cardiovascular disease), their pathways, race-specificity, and overlap with pathways identified in Aim 1. In Aim 3, we apply machine-learning techniques to metabolites to derive a classi- fier for exceptionally healthy aging in Health ABC and apply this prediction rule to young adults in CARDIA to identify young adults at risk for unhealthy aging. We then investigate whether individuals marked for premature aging develop adverse aging phenotypes decades later (neurocognitive, physical, vascular, clinical outcome). This proposal addresses a core mission of the NIA (directly responsive to PA-17-088) by defining a metabolic basis for ?exceptional healthy aging? across race and at the extremes of life. We leverage rich phenotypes within Health ABC, funded efforts within CARDIA, and a team experienced in aging epidemiology, metabolic research, and data science approaches. Successful completion of this project will furnish a rich molecular phe- notypic resource to the aging community for investigation of mechanisms and biomarkers of biological aging.